176 A. L. DAY MINERAL RELATIONS FROM LABORATORY VIEWPOINT 



(1) the effect of the foreign minerals known to be present which directly 

 affected the relation which he sought to establish; (2) the somewhat un- 

 satisfactory character of his temperature and calorimetric measurements. 

 Then, too, the Yan't Hoff-Eaoult relation owes its derivation to the gas 

 laws, and has been found to hold by analogy for aqueous solutions up to 

 concentrations of 1 per cent, but not for any others. It is therefore a 

 serious question whether the analogy can be strained to cover the most 

 concentrated and complicated silicate solutions as A^ogt has tried to make 

 it. Furthermore, none of the constants of A^an't Hoff's law except the 

 temperature have been determined for definite silicates. For these reasons 

 the determinations of the eutectic percentages of orthoclase and quartz, 

 of anorthite and diopside, of diopside and enstatite, and so on through a 

 considerable list, are not satisfactory. Except for the quantitative rela- 

 tions, for which he hardly possessed adequate data, A^ogt has shown in a 

 most comprehensive and suggestive way that the laws of solution apply 

 directly and very generally to silicate solutions, and that the relations of 

 the minerals in the rocks can be determined if the necessary measurements 

 can be made at the temperatures where the formations occur. 



Summary and Conclusion — The geologic Thermometer 



Briefly reviewed, the laboratory situation, then, is this : True character- 

 istics of mineral types can not generally be obtained from natural speci- 

 mens containing the usual quantity of foreign admixtures. If we are to 

 triangulate our field by the use of mineral types, which would appear to 

 be necessary in whatever light their relations be viewed, we must, there- 

 fore, use pure minerals for these types. It has been sufficiently demon- 

 strated that such pure t^^pes can be prepared, and with them all the neces- 

 sary physical characteristics determined. In the discussion of mineral 

 melting temperatures, we differentiate two classes of minerals: (1) those 

 which can be depended on to give a definite point of change of state inde- 

 pendent of the experimental conditions which are provided (that is, which 

 remain in equilibrium throughout), and (2) a group of minerals in which 

 the molecular deorientation corresponding to the change of state is so 

 much hindered by viscosity or similar retarding influences that there 

 exists no particular temperature at which the change will be found to 

 begin, and no particular range of temperature or time interval within 

 which it will proceed to completion. This is a property of this group of 

 substances, and not a limitation of the methods of examining them. The- 

 oretically there is a temperature limit up to which the solid is stable, and 

 above which melting will begin and proceed indefinitely, if sufficient time 

 is given, Conversely, there is a theoretical upper limit to the beginning 



